JP7407098B2 - Tire load prediction system, tire load prediction program, and tire load prediction method - Google Patents

Description

The present invention relates to a tire load prediction system, a tire load prediction program, and a tire load prediction method for predicting the load of a pneumatic tire.

Conventionally, the time-series waveform of an acceleration sensor placed inside a pneumatic tire (hereinafter referred to as a tire) has been used to calculate the tire's contact time ratio (CTR: a parameter indicating the length of contact). A technique has been proposed for estimating the tire load based on the above (Patent Document 1).

Japanese Patent Application Publication No. 2017-161477

However, in the conventional technology, the ground contact length of the tire changes depending on the wear state of the tire, so there is a problem that the error in load prediction is large.

Furthermore, since the centrifugal force of the acceleration sensor is small in a low speed range, it is difficult to make predictions for vehicles that mainly drive at low speeds.

Therefore, the present invention has been made in view of the above problems, and provides a tire load prediction system and a tire load prediction system that can improve the accuracy of load prediction and can perform load prediction even for vehicles that mainly drive at low speeds. The purpose is to provide a program and tire load prediction method.

A tire load prediction system according to one aspect of the present invention includes a sensor unit that is provided inside a tire and has a strain sensor that detects strain in the tire, and acquires strain data in a tire tangential direction output from the sensor unit. a linear transformation unit that linearly transforms the acquired strain data; a first estimation unit that estimates tire tangential direction velocity and angular velocity based on the transformation result by the linear transformation unit; a second estimator that estimates the tire radial speed from the tire tangential speed and angular velocity; and a second estimator that estimates the tire radial speed from the tire tangential speed and angular velocity; and a third estimation unit that estimates a speed in the θ angular direction corresponding to the angle, and estimates a deflection of the tire based on the speed in the tangential direction of the tire, the angular velocity, the speed in the tire radial direction, and the speed in the θ angular direction. The gist of the present invention is to include a fourth estimation section and a load prediction section that predicts a load applied to the tire based on the estimated deflection of the tire.

According to such a configuration, the accuracy of load prediction can be improved.

Further, the fourth estimating unit estimates a deformation profile of the tire based on the speed in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angle direction, and estimates a deformation profile of the tire corresponding to the deflection of the tire. The amount can be extracted to estimate the deflection of the tire.

Thereby, the accuracy of load prediction can be further improved.

In addition, the linear conversion section is configured using the following formula

However, the distortion data can be linearly transformed using ω ₀ : average angular velocity a ₁ : angular velocity magnification ε: distortion. This makes it possible to achieve highly accurate load prediction.

Further, the first estimator calculates the following equation regarding the acquired data:

However, the tangential velocity can be estimated by performing linear transformation using v _T0 : average tangential velocity a ₂ : tangential velocity magnification. This makes it possible to achieve highly accurate load prediction.

A tire load prediction program according to another aspect of the present invention includes a strain data acquisition step of acquiring strain data in a tire tangential direction output from a strain sensor provided on the inner surface or inside of the tire; a linear transformation step for performing linear transformation; a first estimation step for estimating the tire tangential velocity and angular velocity based on the transformation results; a second estimation step of estimating the speed; and a third estimation step of estimating the speed in the θ angular direction corresponding to the angle of the strain sensor with respect to the ground contact surface of the tire, based on the estimated tire tangential acceleration and angular velocity values. an estimating step; a fourth estimating step of estimating the deflection of the tire based on the tire tangential speed and angular velocity, the tire radial speed, and the θ angular direction speed; and a load prediction step of predicting the load applied to the tire based on the tire load prediction system, and is executed by a CPU included in the tire load prediction system.

Thereby, the accuracy of load prediction can be improved.

Further, in the fourth estimation step, a deformation profile of the tire is estimated based on the speed in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angle direction, and a characteristic corresponding to the deflection of the tire is estimated. The amount can be extracted to estimate the deflection of the tire.

Thereby, the accuracy of load prediction can be further improved.

A tire load prediction method according to another aspect of the present invention includes a strain data acquisition process of acquiring strain data in the tire tangential direction output from a strain sensor provided on the inner surface or inside of the tire; A linear conversion process for linear conversion; a first estimation process for estimating the tire tangential speed and angular velocity based on the conversion results; a second estimation step of estimating the speed; and a third estimation step of estimating the speed in the θ angular direction corresponding to the angle of the strain sensor with respect to the ground contact surface of the tire, based on the estimated tire tangential acceleration and angular velocity values. an estimation step; a fourth estimation step of estimating the deflection of the tire based on the tire tangential velocity and angular velocity, the tire radial velocity, and the θ angular velocity; and the estimated tire deflection. and a load prediction process of predicting a load applied to the tire based on the load prediction process.

Thereby, the accuracy of load prediction can be improved.

The fourth estimation step may include estimating a deformation profile of the tire based on the speed in the tangential direction of the tire, the speed in the radial direction of the tire, and the speed in the θ angle direction, The amount can be extracted to estimate the deflection of the tire.

Thereby, the accuracy of load prediction can be further improved.

According to the present invention, it is possible to provide a tire load prediction system, a tire load prediction program, and a tire load prediction method that can improve the accuracy of load prediction and can perform load prediction even for vehicles that mainly drive at low speeds. can.

BRIEF DESCRIPTION OF THE DRAWINGS It is a schematic block diagram which shows the schematic structure of the tire load prediction system based on embodiment. FIG. 1 is a functional block diagram showing a functional configuration of a tire load prediction system according to an embodiment. FIGS. 2A to 2D are explanatory diagrams (a) to (d) showing definitions of coordinate systems used in the explanation of the tire load prediction system according to the embodiment. It is a flow chart which shows the processing procedure of tire load prediction processing performed by the tire load prediction system concerning an embodiment. It is a graph showing the relationship between tangential distortion and time. It is a graph showing the relationship between angular velocity and time. It is a graph showing the relationship between tangential velocity and time. FIGS. 3A to 4C are explanatory diagrams used to explain the principle of estimating tangential velocity. FIGS. It is a graph showing the relationship between sensor angle and time. It is a graph showing the relationship between X-coordinate speed and time. It is a graph showing the relationship between Z coordinate speed and time. It is a graph showing the relationship between X-coordinate displacement and time. It is a graph showing the relationship between Z coordinate displacement and time. It is a graph which shows the situation of tire deflection estimation. It is a graph showing the relationship between tire deflection rate estimation and load.

A tire load prediction system S1 according to an embodiment of the present invention will be described with reference to FIGS. 1 and 2.

In addition, in the description of the following drawings, the same or similar parts are given the same or similar symbols. However, it should be noted that the drawings are schematic and the ratio of each dimension may differ from the actual one.

Therefore, specific dimensions etc. should be determined with reference to the following explanation. Furthermore, it goes without saying that the drawings include portions with different dimensional relationships and ratios.

(Schematic configuration of tire load prediction system)
A schematic configuration of a tire load prediction system S1 according to an embodiment will be described with reference to the schematic configuration diagram of FIG. 1.

Here, the tire load prediction system S1 according to the present embodiment uses a strain sensor SN instead of a conventional acceleration sensor, estimates the tire profile and deflection from the data in one axis (circumferential direction), and calculates the load. This is based on a prediction algorithm (details will be described later).

The tire load prediction system S1 includes a sensor unit SU provided on a pneumatic tire (hereinafter simply referred to as a tire) 10, and a processing device (on-vehicle device) that processes information acquired from the sensor unit SU via a wireless line N1. (ECU) etc.) 200.

FIG. 1 shows a cross-sectional shape of a tire 10 assembled to a rim wheel 90 along the tire width direction.

Further, the tread portion 20 is a portion that comes into contact with a road surface when the tire 10 mounted on a vehicle (not shown) rolls on the road surface. The tread portion 20 is formed with a tread pattern depending on the type of vehicle and required performance.

A sensor unit SU including a uniaxial strain sensor SN for detecting strain in the tire 10 is provided on the inner surface 10a of the tire 10 to which the tire load prediction system S1 can be applied.

Although not directly related to this embodiment, the sensor unit SU may be configured to be able to acquire temperature information and the like in addition to distortion.

In the configuration example shown in FIG. 1, the sensor unit SU is provided on the inner surface 10a facing the tread portion 20. More specifically, the sensor unit SU is attached to the surface of an inner liner (not shown) that prevents leakage of gas such as air filled in the internal space of the pneumatic tire 10 assembled to the rim wheel 90.

It is preferable that the sensor unit SU is provided in each tire 10 mounted on the vehicle. This is because it is desirable to monitor the load on each tire 10 to ensure vehicle safety.

Further, the sensor unit SU does not necessarily have to be attached to the inner surface of the tire 10, and for example, a part or all of the sensor unit SU may be buried inside the tire 10.

(Functional configuration of tire load prediction system)
As shown in the functional block diagram of FIG. 2, the sensor unit SU includes a uniaxial strain sensor SN that detects the strain of the tire 10, a transmitter 101 that transmits detected data to the processing device 200, a strain sensor SN and a transmitter. A battery 102 that supplies power to the device 101 is provided.

On the other hand, the processing device 200 includes a strain data acquisition unit 202 that acquires strain data in the tire tangential direction output from the sensor unit SU via the communication unit 201.

It also includes a storage unit 203 made up of a nonvolatile memory or the like that stores the acquired distortion data.

It also includes a linear conversion section 251 configured with a CPU 250 and the like that linearly converts the acquired distortion data.

Furthermore, a first estimator 252 is provided that estimates the tire tangential direction velocity and angular velocity based on the conversion result by the linear converter 251 .

The second estimator 253 also includes a second estimator 253 that estimates the speed in the tire radial direction from the estimated tire tangential speed and angular velocity values.

The third estimator 254 also includes a third estimator 254 that estimates the speed in the θ angle direction corresponding to the angle of the strain sensor SN with respect to the ground contact surface of the tire 10 from the estimated tire tangential direction acceleration and angular speed.

The fourth estimator 255 also includes a fourth estimator 255 that estimates the deflection of the tire 10 based on the tire tangential velocity and angular velocity, the tire radial velocity, and the θ angular velocity.

Furthermore, it includes a load prediction unit 256 that predicts the load applied to the tire 10 based on the estimated deflection of the tire 10.

Note that the linear conversion section 251, the first estimating section 252, the second estimating section 253, the third estimating section 254, the fourth estimating section 255, and the load predicting section 256 are connected to the CPU 250 and the OS ( This can be realized through cooperation between the operating system (operating system) and a predetermined application program.

Further, details of the processing performed by the linear conversion section 251, the first estimating section 252, the second estimating section 253, the third estimating section 254, the fourth estimating section 255, and the load predicting section 256 will be described later.

(About tire load prediction processing)
With reference to FIGS. 3 to 15, the processing procedure and the like of the tire load prediction process executed by the tire load prediction system S1 will be described.

Before explaining the processing procedure of the tire load prediction process, the definition of the coordinate system used in the tire load prediction system S1 according to the present embodiment is shown in FIGS. 3(a) to 3(d).

Here, R is a radial direction, and T is a tangential direction or a circumferential direction.

Further, the R, T coordinate system is an object coordinate system (local coordinate system) fixed to the strain sensor SN.

On the other hand, x and z are a global coordinate system fixed in space. Further, a represents acceleration, v represents velocity, and u represents displacement. Furthermore, ω represents the angular velocity of the strain sensor SN, and θ represents the angle of the strain sensor SN.

FIG. 4 is a flowchart showing the processing procedure of the tire load prediction process executed by the tire load prediction system S1.

When this process is started, in step S10, strain data in the tire tangential direction is acquired, and the process moves to step S11.

In step S11, the acquired distortion data is linearly transformed and the process moves to step S12.

In step S12, a first estimation process is performed to estimate the tire tangential speed and angular velocity, and then the process moves to step S13.

In step S13, a second estimation process for estimating the speed in the tire radial direction is performed based on the speed in the tire tangential direction and the angular velocity, and the process moves to step S14.

In step S14, a third estimation process for estimating the velocity in the θ angle direction is executed, and the process moves to step S15.

A fourth estimation process for estimating the deflection of the tire 10 is executed based on the speed in the tire tangential direction, the angular speed, the speed in the tire radial direction, and the speed in the θ angular direction, and the process moves to step S16.

In step S16, the load applied to the tire 10 is predicted based on the deflection of the tire 10, and the process ends.

(About load estimation method using biaxial acceleration)
Here, a method for estimating a load using biaxial acceleration related to a comparison target will be briefly described.

First, physically, calculations that follow the two-dimensional motion of an object (rigid body) to find a trajectory have three degrees of freedom, and in order to estimate the tire load using two-axis acceleration, "R Data on three axes are required: "direction", "T direction", and "θ direction".

Therefore, the trajectory inside the tire is assumed to be mostly circular motion, the acceleration in the two axes (R, T) directions is measured, the angular velocity is estimated by calculation, and then the trajectory is determined.

As a result of intensive research, the present inventor has determined how to calculate the three axes ("R direction", "T direction", "θ direction") from the one axis measurement data in order to realize the above processing with the one axis strain sensor SN. We have developed a new method (algorithm) for estimating the data of

(About the load prediction method in the present invention)
In the load prediction method according to the present invention, first, measured strain data is linearly transformed to estimate values of T-direction velocity and angular velocity.

Next, the R direction acceleration is estimated from these two estimated values. Then, by integrating the R-direction acceleration to the R-direction velocity, it is possible to obtain data for three axes: "R-direction velocity," "T-direction velocity," and "θ-direction angular velocity." By using these data, it becomes possible to calculate the trajectory.

The details of the calculation method will be explained step by step below.

Note that for convenience of explanation, sample data shown in FIG. 5 will be used here. This is data obtained by FEM analysis, and the tire type is "11R22.5M801."

Moreover, this data is strain data on the inner surface of the tire measured by the strain sensor SN, and is hereinafter expressed as ε(t).

[Estimation of angular velocity]
First, the process of calculating angular velocity from tangential strain will be explained.

The angular velocity is estimated by linearly transforming the measured data using the following equation (Equation 1).

However, ω ₀ and a ₁ are parameters indicating the average angular velocity and the angular velocity magnification, respectively.

Note that ω ₀ is determined by the following formula (Equation 2) using one rotation time (ORT: One Rotational Time) determined from the rotation period of the tire.

Further, _a1 is determined by learning from experimental values or by trial and error. Note that nonlinear transformation may be used.

FIG. 6 shows a graph of the estimated angular velocity.

[Angular velocity estimation principle]
In the tire 10, large bending deformation occurs in a region near the ground contact edge, and large compressive strain occurs on the inner surface of the tire.

Further, it is considered that the curvature increases due to large bending deformation, and the angular velocity of the strain sensor SN itself increases.

On the contrary, in the vicinity directly below the axle, bending deformation is relaxed, compressive strain is reduced (becomes tensile force), and the angular velocity of the strain sensor SN itself is considered to be reduced due to the reduced curvature.

From this, it is assumed that there is a correlation between the compressive strain and the angular velocity of the strain sensor SN, and that it is linear here.

[Estimation of tangential velocity]
Next, a process of calculating tangential velocity from tangential strain will be explained.

The tangential velocity is estimated by performing linear transformation on the measured data using the following equation (Equation 3).

However, v _T0 and a ₂ are parameters indicating the average tangential velocity and the tangential velocity magnification, respectively.

Note that v _T0 is determined by the following formula (Equation 4) using one rotation time (ORT: One Rotational Time) determined from the rotation period of the tire and the tire radius R.

Further, _a2 is determined by learning from experimental values or by trial and error. Note that nonlinear transformation may be used.

FIG. 7 shows a graph of the estimated tangential velocity.

"Principle of estimating tangential velocity"
Assume that the rolling tire 10 is divided into spaces at minute sampling time intervals Δt, as shown in FIG. 8(a).

Here, when the tire 10 is in the state before deformation (the tire is floating in the air and rotating), assuming a time-independent radius R ₀ and an angular velocity ω ₀ , the distance traveled in a minute time is R ₀ ω ₀ △t (See Figure 8(b)).

On the other hand, when the tire 10 is in contact with the ground and is deformed, the radius and angular velocity are functions of time, and are expressed as R(t) and θ'(t), respectively. At this time, the distance traveled in a minute time can be written as R(t)θ'(t)Δt (see FIG. 8(c)).

Since distortion is defined as the extent to which it has expanded or contracted relative to its original length, the circumferential distortion of the tire 10 is expressed by the following equation (Equation 5).

Here, assuming that the strain sensor SN is in circular motion, the tangential velocity v _T (t) can be written by the following equation (Equation 6).

By substituting this into the expression for ε(t) in Equation 6, Equation 7 is obtained.

Here, since R ₀ ω ₀ is a fixed value that does not depend on time, it can be seen that v _T (t) and ε(t) are in a linear transformation relationship.

[Estimation of radial acceleration/velocity]
Assuming that the strain sensor SN is moving in a circular motion, it can be theoretically estimated by calculating the following equation (Equation 8).

Further, the velocity in the radial direction is calculated by time-integrating the estimated value using the following equation (Equation 9).

However, a _R (t) may involve preprocessing such as centering in advance. Further, v _R0 is set to an arbitrary value as an initial value.

[Coordinate transformation]
Here, the R, T coordinate system is transformed into the X, Z coordinate system using the estimated radial velocity v _R (t) and the angular velocity ω(t), which is the estimated tangential velocity v _T (t). .

First, the angular velocity ω(t) is integrated, and the angle θ(t) is calculated using the following equation (Equation 10).

However, preprocessing such as adjusting ω(t) so that the average becomes 2π may be performed. The initial value θ ₀ is set to an appropriate value such as 0 or -π.

The graph of FIG. 9 shows an example of the calculation result of the angle θ(t).

Next, coordinate transformation of velocity is performed.

Calculation is performed from the object coordinate system (R, T) of the strain sensor SN to the global coordinate system (x, z) using the following equation (Equation 11).

Examples of the calculated results are shown in the graphs of FIGS. 10 and 11.

[Estimation of displacement]
The velocity is integrated and the displacements u _x (t) and u _z (t) are calculated from the following formula (Equation 12).

However, v _x and v _z may require preprocessing such as centering. The initial values u _x0 and u _z0 can be set to arbitrary values. Examples of the calculated results are shown in the graphs of FIGS. 12 and 13.

[Profile and deflection estimation]
Then, when the displacement (u _x (t), u _z (t)) excluding the time component is plotted on a two-dimensional plane, it becomes a trajectory shown in Fig. 14.A circle is fitted to this trajectory to calculate the regression radius R. Find _fit .

Furthermore, the effective radius R _eff (for example, the minimum value from the center of the regression circle) when the tire is bent is determined.

Then, the deflection is calculated using the following equation (Equation 13).

Note that the feature amount used for load prediction may be either the deflection d or the deflection rate d/R _fit .

[Load prediction]
Next, the effectiveness of load calculation using deflection and its method will be shown.

First, FIG. 15 shows the relationship between the estimated deflection rate calculated using the algorithm of the present invention and the load. Note that this data is also an example of analysis performed using FEM.

Further, in FIG. 15, the plot 10A shows a normally worn tire, and the plot 10B shows a tire whose center part is unevenly worn.

In this way, a linear relationship was obtained, similar to the deflection determined from the axle height described above.

Moreover, as shown in FIG. 15, it can be seen that the plots 10A and 10B almost converge to one line.

As described above, since the dependence on wear, which is a disadvantage of CTR, is not observed, it is possible to improve the accuracy of load prediction.

Furthermore, when actually predicting the load, the load is calculated using the following formula (Equation 14).

In addition to the deflection information, TPMS information and wear information can also be used as explanatory variables, and the parameters can be identified through machine learning in advance before operation.

Furthermore, although linear multiple regression is basically sufficient, a nonlinear model may also be used.

As described above, according to the tire load prediction system S1 according to the present embodiment, the accuracy of load prediction can be improved, and load prediction can be performed even for vehicles that mainly drive at low speeds.

Further, by using the strain sensor SN, tire deflection and load can be predicted regardless of the wear state of the tire.

Although the tire load prediction system and tire load prediction program of the present invention have been described above based on the illustrated embodiments, the present invention is not limited thereto, and the configuration of each part can be modified using any arbitrary structure having similar functions. It can be replaced with the configuration.

For example, if operating conditions such as power supply (battery) 102 are satisfied, part of the processing functions of processing device 200 in this embodiment may be installed in sensor unit SU.

S1 Tire load prediction system SU Sensor unit SN Strain sensor 10 Pneumatic tire (tire)
200 Processing device 202 Data acquisition unit 203 Storage unit 250 CPU
251 Linear conversion unit 252 First estimation unit 253 Second estimation unit 254 Third estimation unit 255 Fourth estimation unit 256 Load prediction unit

Claims (8)

a sensor unit provided inside a tire and having a strain sensor that detects distortion of the tire;
a strain data acquisition unit that acquires strain data in a tire tangential direction output from the sensor unit;
a linear transformation unit that linearly transforms the acquired distortion data;
a first estimation unit that estimates a tire tangential direction velocity and an angular velocity based on the conversion result by the linear conversion unit;
a second estimation unit that estimates the speed in the tire radial direction from the estimated tire tangential direction speed and angular velocity;
a third estimation unit that estimates a velocity in the θ angular direction corresponding to the angle of the strain sensor with respect to the ground contact surface of the tire from the estimated tire tangential direction acceleration and angular velocity;
a fourth estimation unit that estimates the deflection of the tire based on the speed and angular velocity in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angular direction;
a load prediction unit that predicts a load to be applied to the tire based on the estimated deflection of the tire;
Tire load prediction system. The fourth estimation unit estimates a deformation profile of the tire based on the speed in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angle direction, and calculates a feature amount corresponding to the deflection of the tire. The tire load prediction system according to claim 1, wherein the tire load prediction system extracts and estimates the deflection of the tire. The linear conversion section has the following formula

The tire load prediction system according to claim 1, wherein the distortion data is linearly transformed using ω ₀ : average angular velocity a ₁ : angular velocity magnification ε: strain. The first estimator calculates the following equation regarding the acquired data:

The tire load prediction system according to claim 1, wherein the tangential speed is estimated by linear transformation using v _T0 : average tangential speed a ₂ : tangential speed multiplier. a strain data acquisition step of acquiring strain data in a tire tangential direction output from a strain sensor provided on the inner surface or inside of the tire;
a linear transformation step of linearly transforming the acquired strain data;
a first estimation step of estimating the tire tangential speed and angular velocity based on the conversion result;
a second estimation step of estimating the speed in the tire radial direction based on the estimated tire tangential direction speed and angular velocity;
a third estimation step of estimating a velocity in the θ angular direction corresponding to the angle of the strain sensor with respect to the ground contact surface of the tire, based on the estimated tire tangential direction acceleration and angular velocity;
a fourth estimation step of estimating the deflection of the tire based on the tire tangential velocity and angular velocity, the tire radial velocity, and the θ angular velocity;
a load prediction step of predicting a load to be applied to the tire based on the estimated deflection of the tire;
A tire load prediction program that is executed by a CPU included in a tire load prediction system. The fourth estimation step estimates a deformation profile of the tire based on the speed in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angle direction, and calculates a feature amount corresponding to the deflection of the tire. The tire load prediction program according to claim 5, wherein the tire load prediction program extracts and estimates the deflection of the tire. a strain data acquisition step of acquiring strain data in a tire tangential direction output from a strain sensor provided on the inner surface or inside of the tire;
a linear transformation process for linearly transforming the acquired strain data;
a first estimation step of estimating the tire tangential velocity and angular velocity based on the conversion result;
a second estimation step of estimating the speed in the tire radial direction based on the estimated tire tangential speed and angular velocity;
a third estimation step of estimating the velocity in the θ angular direction corresponding to the angle of the strain sensor with respect to the ground contact surface of the tire, based on the estimated tire tangential direction acceleration and angular velocity;
a fourth estimation step of estimating the deflection of the tire based on the tire tangential velocity and angular velocity, the tire radial velocity, and the θ angular velocity;
a load prediction step of predicting a load to be applied to the tire based on the estimated deflection of the tire;
A tire load prediction method having the following. In the fourth estimation step, a deformation profile of the tire is estimated based on the speed in the tire tangential direction, the speed in the tire radial direction, and the speed in the θ angle direction, and a feature amount corresponding to the deflection of the tire is estimated. The tire load prediction method according to claim 7, wherein the tire load prediction method is extracted and the deflection of the tire is estimated.

Images